Project Summary/Abstract Schizophrenia (SCZD) is a severe mental disorder that imposes a significant burden on public health, and though the disorder is highly heritable, the nature of the genetic contribution is poorly understood. Recent efforts in combining existing genome-wide association studies (GWAS) of SCZD by the Psychiatric Genomics Consortium (PGC) have led to the strongest credible reports of genetic associations with the disorder. However, the neurobiological mechanisms by which the implicated variants increase the risk for SCZD are unknown. Here we will expand upon the rich genomic data generated from the Lieber Institute for Brain Development (LIBD) by incorporating genome-scale DNA methylation (DNAm) data on the same carefully characterized subjects that have RNA sequencing (RNA-seq) and genetic data to better determine how genetic risk for SCZD manifests in the human brain. We will utilize new experimental approaches that can quantify both methyl-cytosine (5mC) and hydroxymethyl-cytosine (5hmC) to untangle total methylation signal present in previous smaller studies using whole genome bisulfite sequencing (WGBS). In this proposal we will perform whole genome bisulfite sequencing on 600 samples (300 donors across 2 brain regions) and correlate the resulting DNAm levels with genotype, diagnosis, and local expression levels to better understand epigenetic regulation of schizophrenia risk in the frontal cortex and hippocampus. We will perform methylation. We will first perform methylation quantitative trait loci (meQTL) analysis with genetic risk variants for SCZD identified in the PGC, as well as all common variants in these samples, separately for 5mC and 5hmC levels to determine potential epigenetic mechanisms underlying risk, and hypothesize increased statistical power by decomposing total DNAm signal. We will then identify genome- wide significant differentially methylated regions (DMRs) comparing patients with schizophrenia to matched non-psychiatric controls within and across brain regions using 5mC and 5hmC across both CpG and non-CpG sites. These regions can then be interrogated for potential functionality by correlating DNAm levels within DMRs to the matched expression data via RN-seq. Lastly, we will identify functional correlates of 5mC and 5hmC DNAm levels by combining WGBS and RNA-seq on the same samples across the entire methylome and transcriptome, both within and across diagnostic groups - by further combining genetic data, we can identify the subset of DNAm-expression correlations driven by genetic versus epigenetic variation. Proximal cellular phenotypes like DNAm levels may ultimately show strong and meaningful association with risk alleles that can further mediate gene expression levels. In the grant, we aim to elucidate some of the molecular biology underlying genetic risk and molecular signatures of schizophrenia and thereby help identify novel targets for intervention in the disease process and potential treatment strategies.